Route Optimization and Internet of Things (IOT)-Enabled Smart Waste Management System in Urban Areas: A Case Study of Rivers State, Nigeria

Abstract

This paper presents Internet of Things (IoT)-enabled smart waste management system with integrated route optimisation, aimed at addressing the challenge of inefficient logistics planning and resource control in urban waste management. It leverages IoT technologies including ultrasonic sensors for fill-level detection, GIS and GPS modules for real-time location tracking and shortest path measurements and Arduino microcontrollers for system wide data acquisition from sensors enabling real-time monitoring of waste bins across Port Harcourt and major city centers in Rivers State, Nigeria. The data collected from the sensors is transmitted via lightweight protocols (MQTT/HTTP) to a cloud database, where it is processed and visualised through a Flask-powered dashboard. The system is designed using the Object-Oriented Design (OOD) methodology, which allows modular representation of system components such as bins, sensors, and route optimisation and recommendation logic. The route optimisation approach entails the usage of a Shortest Path Algorithm combined with a Route Recommender Model implemented to minimize travel distance, prioritize bins nearing full capacity, and avoid redundant collection trips. The recommender model was trained on simulated bin data (location, fill-level, collection frequency) using Decision Trees. Simulation results show that the scores were high across all metrics, with Precision at 0.89, Recall at 0.87, F1-Score at 0.88, and 92.5% accuracy in predicting optimal collection and route optimisation completion rate of 35%, which indicates that the results of the proposed IoT-enabled system significantly reduce operational costs, travel distance, and energy consumption while improving waste collection efficiency, and promoting sustainable urban sanitation practices. These values also indicate that the model not only predicts bin status accurately but also maintains a good balance between identifying actual full bins and minimising false alerts.